Would you say that a display and a printer are a perfect painter because they can render images? And a speaker is a very good musician because they can produce sound?

The LLM tasks is to produce a string of words according to an internal model trained on texts written by humans (and now generted by other LLMs). This is not intelligence.